Procedural Drift Causes and Consequences

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

Procedural Drift: Causes and Consequences

Robert I. Baron, Ph.D


The Aviation Consulting Group
November, 2017

Procedures, policies, and checklists are an effective way to ensure safety-related tasks are

being conducted in a standardized, pragmatic way. They are also, in most cases, a regulatory

requirement. Thus, all people employed in aviation (whether they are in the cockpit, maintenance

hangar, or the ramp) are following the procedures…right?

Wrong. Procedural noncompliance, or procedural drift, has been either a primary, or

contributing, causal factor in the majority of aviation accidents. The term procedural drift refers

to the continuum between textbook compliance, and how the procedure is being done in the real

world. Procedural drift is not something new. We have been drifting for a very, very, long time.

However, recent accidents have illuminated the ubiquity and severity of the problem, prompting

the U.S. National Transportation Safety Board (NTSB) to add ‘Strengthen Procedural

Compliance’ to its Most Wanted List of Aviation Safety Improvements (2015). This NTSB

recommendation was largely in response to an Execuflight Hawker HS125-700A crash on

approach to Akron, Ohio, in 2015. All onboard perished (2 crew and 7 passengers). For the

author’s analysis of this accident from the procedural drift in flight operations perspective, you

can visit tacgworldwide.com/Failure-to-Follow-Procedures. The rest of this paper, however, will

focus on procedural drift as it applies to aviation maintenance, although the principles can be

applied to any segment of the aviation industry.


Deviations from approved procedures continue to be a leading cause of maintenance-

related aircraft accidents. But why is it so hard to just follow the procedures? The answer to this

question can range from very simple to quite complex (refer to the diagram below).

Reasons for not Following Procedures. Copyright ©2017, Robert Baron

Procedure(s) not Available

There are documented procedures for just about all maintenance tasks. However, there

are cases where either a documented procedure does not exist, or the procedure exists but is not

available. In either case, the question must be asked (preferably proactively rather than

reactively); why is the procedure not available? Is there even a written procedure for the task?

There are many cases where an aircraft maintenance technician (AMT) must conduct a task
when no written procedures are available from the manufacturer. This can lead to knowledge-

based errors, which can result from the AMT using “head knowledge” only, as he/she attempts to

conduct the task solely from past experiences, or familiarity by having done a “similar” task.

This can lead to flawed assumptions, and consequently, a maintenance error.

On the other hand, procedures may be readily available, but due to inadequate

dissemination of information, they are not finding their way onto the AMT’s computer screen.

This could be indicative of an organizational problem. The solution for this lack of

communication should not focus on ascribing blame to management or an individual, but rather

to investigate the “how and why” of the problem and then work collaboratively to correct it.

Procedure(s) Available but Inadequate (i.e., incomplete or incorrect)

In this situation, there is no problem with procedure availability, however the procedure

itself may be erroneous or flawed (i.e., incomplete or incorrect). If the procedure is inadequate,

the AMT might follow the procedure, only to find that an error has been made in the execution

of the task.

In 2003, a Colgan Air Beechcraft 1900D, with just the pilots aboard (who were fatally

injured), crashed shortly after takeoff due to a reversal of the elevator trim system. Although

there were a number of links in the causal chain, one of those links included a pictorial error in

the aircraft maintenance manual. Specifically, the elevator trim drum graphic was incorrectly

depicted, which if followed, would cause a reversal of the action of the elevator manual trim

system. That is exactly what happened. Compounding matters further, a functional check, which

would have detected the error, was not accomplished.

Some studies have been conducted in the area of maintenance documentation quality.

NASA researcher Barbara Kanki and her colleagues found that procedural errors, which are
defined as any information-related error involving documents, have been implicated in 44% to

73% of maintenance errors. The three most problematic areas identified were inspection and

verification issues (34%), incompleteness of the documents (27%), and incorrectness of the

documents (22%). Similarly, in a study of 458 ASRS reports submitted by AMTs, Kanki and her

colleagues found that the most frequently cited maintenance document deficiencies were missing

information (48%), incorrect information (19%), difficult to interpret (19%), and conflicting

information (19%). Another study, by Wichita State University researcher Bonnie Rogers and

her colleagues, investigated Publication Change Requests (PCRs) by AMTs on the types of

errors found in aircraft maintenance manuals published by four manufacturers. Results showed

that the majority of PCRs related to procedures found in Flight Controls, Landing Gear, and

Powerplant systems. The highest percentage of PCRs involved Procedural Errors (42.5%)

followed by Language (29.9%), Technical (16.5%), Graphic (8.1%), and Effectivity (n/a).

Common procedural errors were categorized as Step(s), Ordering, Alternate method,

Check/Test/Inspection, Caution/Warning. Language errors included typographical errors

(Typos), grammatical errors (Grammar), a need for clarification of the information (Clarity), and

inaccurate information within a step (Incorrect).

Clearly, manufacturers’ maintenance documentation may, in and of itself, be part of the

overall problem related to procedural drift. However, these documentation deficiencies should

not be viewed as a shift of blame from the AMT to the manufacturer. Until documentation issues

such as missing information, incorrect information, difficulty in interpretation, and conflicting

information are resolved, it will be difficult to sell the idea of following approved procedures to

AMTs on a far-reaching basis. Manufacturers and AMTs need work together on this. AMTs

should continue to identify and provide PCRs to manufacturers, and manufacturers should
realize that their documentation can be a significant contributor to procedural drift. Keep in mind

that many of the procedures in a maintenance manual have been written by engineers who may

have never performed the procedure on an actual aircraft. The result is a procedure that may be

unrealistically transferable to the practical working environment. This in turn could account for

an AMT making up his/her own way of performing the procedure.

AMT has a “Better Way” of Performing the Procedure

In this case, a written procedure may be available, but the AMT performs a task in a way

that is different from the steps outlined in the procedure. In many cases, the AMT feels that

he/she has “a better way of performing the procedure,” including skipping steps that the AMT

may deem “unnecessary.”

In 2003, an Air Midwest Beechcraft 1900D experienced a loss of control on takeoff from

Charlotte-Douglas International Airport (CLT) in Charlotte, North Carolina, U.S. The NTSB

investigation revealed that overloading and an aft CG combined with incorrect rigging of the

elevator were causal to the accident. The incorrect rigging was attributable to an AMT skipping

steps in the full rigging procedure because he deemed the steps “unnecessary.” A functional

check was not conducted, nor was it required, at the time of the accident. A functional check

would have detected the maintenance error, and subsequently became a requirement after this

accident. All occupants (2 crew and 19 passengers) received fatal injuries.

Some AMTs are still doing things “their own way” even when the documentation is

considered effective and comprehensible. A major reason for this may include shortcutting to

save time on perceived “unimportant steps.” Other reasons may include complacency and norms.

These are factors that can be somewhat easier to mitigate at the person level as compared to
organizational factors such as pressure, fatigue, and lack of resources, which can require a

paradigmatic shift in the overall safety culture of a company.

Then there is the issue of informal sources of information, such as the “Black Book.”

AMTs will literally rewrite, or modify, procedures and store them in their so-called Black Book.

There are a few reasons for this. One reason is the procedure itself is flawed, incorrect, or just

doesn’t make sense (i.e., Step A cannot be done unless Step D is done first). In this case, it is

understandable that an AMT would need to modify the procedure.

On the other hand, a more ominous reason, and more likely, is that the AMT has figured

out “a better way of getting it done.” Typically, the reason centers on time-efficiency, shortcuts,

or skipping what the AMT considers “unnecessary steps.” In any case, the procedures have been

written into the AMTs informal Black Book, but of course the AMT will categorically deny any

existence of said book.

The informal source of information doesn’t necessarily have to be in the form of a Black

Book, either. It can also be in the unwritten form of tribal knowledge (procedures that are being

conducted contrary to manufacturers specifications and being informally handed down from

“generation to generation” of AMTs).

Procedure is Adequate but AMT Doesn’t use it

This situation can be most problematic. These types of violations can fall under two

broad categories; routine and situational. A routine violation (also known as a norm), happens at

a collective level. “If everyone else is doing it, then why shouldn’t I?” An example of this would

be AMTs not conducting a borescope inspection because “it never leaks.” This type of mindset

can spread quickly, and is sometimes referred to as a normalization of deviance. On the other

hand, a situational violation is typically a one-off, where an AMT does not strictly adhere to a
procedure because of an unpropitious working condition (i.e., time pressure, fatigue, or

availability of resources).

In 2007, an American Airlines MD-82 experienced an in-flight engine fire requiring a

turnback and emergency landing in St. Louis (STL). The NTSB investigation revealed that a

component in the manual start mechanism of the engine was damaged when a mechanic used an

unapproved tool to initiate the start of the #1 (left) engine while the aircraft was parked at the

gate at STL. The deformed mechanism led to a sequence of events that resulted in the engine

fire, to which the flight crew was alerted shortly after take-off. There were no injuries or

fatalities.

The above is an example of a typical situational violation. The term “violation” applies

because the AMT has full volition of his/her actions and thus full accountability. This contrasts

with no conscious awareness of committing an error, such as forgetting to put an oil cap back on

an engine. From a disciplinary standpoint, these distinctions are important.

Conclusion

From the organizational level, an effective Safety Management System (SMS) can help

to identify procedural drift through mechanisms such as employee reporting, safety performance

monitoring, and incident/accident investigation. But this is not a magic bullet. Much of the drift

is happening clandestinely, and many of the erroneous outcomes only become known reactively,

as has been exemplified in the accidents/incidents mentioned in this article. The key to reducing

procedural drift is to work proactively to identify the precursors to drift, and then address them

accordingly. This begins with a healthy safety culture, which begins at the top of the

organization.
Management should be very cognizant of its influences on line personnel. Company

leaders may be the driving force that creates that pressure, fatigue, and lack of resources that lead

to errors at the AMT level, such as procedural deviations and violations. Also, keep in mind that

simple exhortations such as “make sure you always do a functional check” will not be effective.

Organizations need to emphasize to their AMTs, in an ongoing and holistic manner, the critical

importance of following approved procedures.

From the AMT level, an awareness of the consequences of intentional procedural

deviations and violations is a great start. This should be covered in human factors courses and

hopefully reinforced on a continuous basis. Even then, many AMTs that succumb to procedural

drift have been taught the importance of strictly following procedures. Many times, these

procedural deviations are well-intentioned with no maleficence intended, and yet they wind up

causing some of the worst maintenance-related accidents. Ask yourself, “if something bad

happens, can I live with it?”

Keep in mind that some of the biggest problem areas are skipping steps, signoffs without

verification, or continuing a job without the correct tools or equipment. Also, I would be remiss

if I did not make special mention of the consequential effects of skipping required

functional/operational checks. Not only is this step a required component of many tasks, but it is

also, in many cases, the last chance to trap any maintenance errors that may have been made in

the job completion process. I estimate that at least half of all maintenance-related accidents

occurred with a corresponding omission of a functional/operational check. Think about that the

next time you decide to skip this check in the interest of time. Those extra few minutes could

save you your job, an aircraft, and even a few lives. Do you catch my drift?
Dr. Robert (Bob) Baron is the President and Chief Consultant of The Aviation Consulting Group
(TACG). His specializations include Human Factors (HF), Safety Management Systems (SMS),
Crew Resource Management (CRM), Line Operations Safety Audit (LOSA), and Fatigue Risk
Management (FRM). He consults with, and provides training to, hundreds of aviation
organizations on a worldwide basis.

You might also like